Abstract

In this paper, the performance of speaker modeling schemes such as vector quantization (VQ) and Gaussian mixture model (GMM) is compared for speaker identification. Along with the effective size of feature set, model based approaches are typically used as a solution for robustness issues of speaker recognition systems. Gaussian Mixture Model (GMM) is versatile parameter estimation approach whereas; Vector Quantization (VQ) is based on template modeling. Here, first, MFCC features are used to extract speaker specific speech features for text-independent speaker identification. MFCC features are then modeled using Vector Quantization (VQ) and Gaussian mixture model (GMM) and their performance is compared in the context of speaker identification. The average recognition rate achieved for MFCC with GMM is 99.2% and for MFCC with VQ is 98.4% on TIMIT database consisting of 64 speakers.

References

• R. Togneri and D. Pullella, “An overview of speaker identification: Accuracy and robustness issues,” IEEE Circuits Syst. Mag., vol. 11, no. 2, pp. 23–61, Second Quarter 2011.
• R.Shantha Selva Kumari, S. Selva Nidhyananthan, Anand.G, “Fused Mel Feature sets based Text-Independent Speaker Identification using Gaussian Mixture Model”, International Conference on Communication Technology and System Design 2011, Journal on Procedia Engineering, Elsevier, vol. 30, pp. 319–326, 2012.
• Mangesh S. Deshpande, Raghunath S. Holambe, “New Filter Structure based Admissible Wavelet Packet Transform for Text-Independent Speaker Identification”, International Journal of Recent Trends in Engineering, vol. 2, no. 5, pp. 121-125, 2009.
• Sumithra Manimegalai Govindan, Prakash Duraisamy, Xiaohui Yuan, “Adaptive wavelet shrinkage for noise robust speaker recognition”, Journal on Digital Signal Processing, Elsevier, vol. 33, pp. 180-190, 2014.
• Noor Almaadeed, Amar Aggoun, Abbes Amira, “Speaker identification using multimodal neural networks and wavelet analysis”, IET Journals and Magazines, vol. 4, no. 1, pp. 18-28, 2015
• Khaled Daqrouq, Tarek A. Tutunji, “Speaker identification using vowels features through a combinedmethod of formants, wavelets, and neural network classifiers”, Journal on Applied Soft Computing, Elsevier, vol. 27, pp. 231-239, 2015.
• M.Hassan Shirali-Shahreza, Sajad Shirali-Shahreza, “Effect of MFCC Normalization on Vector Quantization Based Speaker Identification”, Signal Processing and Information Technology (ISSPIT), 2010, pp.250-253, December 2010.
• Pawan K. Ajmera, Dattatray V. Jadhav, Ragunath S. Holambe, “Text-independent speaker identification using Radon and discrete cosine transforms based features from speech spectrogram”, Journal on Pattern Recognition, Elsevier, Vol.44, Issue 10-11, Pages 2749-2759, 2011.
• Amol A. Chaudhari, S. B. Dhonde, “Effect of Varying MFCC Filters for Speaker Recognition”, International Journal of Computer Applications, vol. 128, no. 14, pp. 7-9, October 2015.
• Tomi Kinnunen, Haizhou Li, “An overview of text-independent speaker recognition: From features to supervectors”, Journal on Speech Communication, Elsevier, vol. 52, no. 1, pp. 12–40, 2010.
• Holambe, Raghunath S., Deshpande, Mangesh S., “Advances in Non-Linear Modeling for Speech Processing”, SpringerBriefs in Speech Technology, Section 2, Section 6, pp. 11-15, 77-82, ISBN 978-1-4614-1505-3, 2012.
• J.S. Garofolo, L.F. Lamel, W.M. Fisher, J.G. Fiscus, D.S. Pallett, N.L. Dahlgren, V. Zue, TIMIT acoustic-phonetic continuous speech corpus, http://catalog.ldc.upenn.edu/ldc93s1, 1993.
• Md Jahangir Alam , Tomi Kinnunen , Patrick Kenny , Pierre Ouellet, Douglas O’Shaughnessy, “Multitaper MFCC and PLP features for speaker verification using i-vectors”, Journal on Speech Communication, Elsevier, vol. 55, no. 2, pp. 237-251, 2013.